Update README.md

README.md

@@ -8,16 +8,18 @@ pinned: false
 short_description: VLM assets for Llama-3.2-11B-Vision-Instruct
 ---
 
-# 🇳🇴🇩🇰 Vision Language Model assets
-This repo is an effort to support open source development of VLM's in the Scandinavian languages.
+# 🇳🇴🇩🇰 Open Source Vision Language Model assets
+Building on the philosophy of open source 🦙, this repo is an effort to support development of small VLM's in the Scandinavian languages. Aa we are only fluenr in Norwegian and Danish, we have focused on these two languages. However, we encourgage the community () to help build on our work and extend the coverage. The current models and data focus on transcription and annotiation of documents in Norwegian and Danish, going beyond the capabilities of OCR.
 
 # In these collections you will find:
 - 💽 Datasets for fine-tuning VLM
   - 🇳🇴 MykMaks/NorwegianDataset-compressed
   - 🇩🇰 MykMaks/fmsudgivelser
 - 💾 Training code
+  - MM checkpoints: https://github.com/mikeriess/llama33_resources/
+  - V-I checkpoints: https://github.com/mikeriess/llama33_resources/
 - 🤖 Model LORA-adapter checkpoints for Llama-3.2-11B-Vision-Instruct
   - The model is iteratively trained over all datasets:
-    - The suffix of each file denotes the order checkpoint, along with the dataset that it was fine-tuned on
+    - The suffix of each file denotes the order of the checkpoint, along with the dataset that it was fine-tuned on
 - 💸 Final merged model:
   - <b>Llama-3.2-11B-Vision-Instruct-MykMaks</b>